The Data Drop: Staff absence in schools

The Data Drop: Staff absence in schools

Welcome to The Data Drop. Our philosophy at Arbor is that staff should have direct, easy access to their data – and be able to do something meaningful with it. Our MIS is built on a world-class data platform, which allows us to deliver things like national-level benchmarking to schools and trusts within seconds, rather than hours or days. And allows schools to make real-time, data-driven decisions. 

This new series is designed to share insights from our data warehouse (containing the data of c.4,000 schools), to shed light on the wider patterns emerging across the education sector.

–

Reasons for staff absence in schools

In a post-Covid world, and also in a year marked by strikes and an emphasis on staff wellbeing, we thought we’d use our first blog in our Data Drop series to take a look at staff absence data. 

When looking at staff absence data based on the last five academic years between 1st September – 24th April (in order to be comparable to the 22/3 dataset at time of publication), there’s a significant downward trend in regards to sickness-related absences in school staff – but only when it comes to absences that have not been certified.

–

–

–

Only 26.8% of recorded staff absences in the academic year 2022-3 were as a result of sickness (certification unknown), dropping from 37.8% the previous year. This is a further jump from pre-Covid; in the academic year 2018-19, the proportion of working days lost attributed to sickness (certification unknown) was 46.7%. 

By contrast, there has been a rise in other types of sickness absence since 2018. This is the case where absence has been medically certified as well as self-certified.

–

Absence due to school strikes

The other absence category which is seeing an unsurprising spike this academic year is absences related to strike action. In 22/23, 2.3% of working days missed were related to strikes. The proportion of working days missed by strikes has sat below 0.1% across the previous four years. 

This equates to 41,193 days lost so far (to 24/04/2023) in 22/23 (across 3,631 many schools from where we drew this dataset), compared with a combined total of 260.5 days across the previous four years. Whilst these stats must also take into account the change in number of Arbor’s schools which make up the data, the difference is still notable. You can see more of Arbor’s data around the strikes in our work with SchoolsWeek here.

–

Other observations:

• Authorised absence: The proportion of working days lost to authorised absence has dropped significantly over the past five years, from its peak at 19.7% in 19/20 to 4.8% in 22/23

• Self-isolation: As can be expected given the dramatic fall in Covid cases, the proportion of absence as a result of self-isolation fell from 7.4% in 21/22 to 0.6% in 22/23

• Sickness types: Sickness due to cold, cough, flu, viral infection has dropped from 18.8% to 15.6% (22/23 vs 21/22)

• Interviews: Absence due to staff attending interviews at other schools is up to more than double pre-pandemic levels (0.5% in 22/23 vs 0.2% in 18/19 and 0.1% in 19/20). This could be due to the high number of vacancies in the sector, giving more agency to teachers in search of better pay and/or opportunities
• Wellbeing: Interestingly, wellbeing leave is up from 0.2% of total leave in 21/22 to 0.3% in 22/23. This could reflect an increased focus on wellbeing from schools, a reduction in overall wellbeing of school staff (leading to an increase in staff needing to take wellbeing days off/emergency wellbeing days), or a recognition from schools that they need to record leave taken specifically for wellbeing, that would have previously been recorded as other forms of leave. It’s also important to note that despite this growth, wellbeing leave still represents a very small proportion of working days lost  
• TOIL: Time off in lieu absences are also up to more than double pre-pandemic levels (0.4% 22/23 vs 0.1% in 18/19 and 19/20). Though these numbers are very low, this could still be a nod to an increase in workload 

–

Methodology 

Periods covered – we’ve chosen to use these dates so that we have a comparable dataset to the 22/23 year at time of writing (May 2023). 

    22-23: 01/09/2022 – 24/04/2023

    21-22: 01/09/2021 – 24/04/2022

    20-21: 01/09/2020 – 24/04/2021

    19-20: 01/09/2019 – 24/04/2020

    18-19: 01/09/2018 – 24/04/2019

Staff absence is recorded using startdate, enddate and working days lost in between. Due to school holidays this is difficult to split days into the correct periods for long-term absence. Startdate has been used as an approximation for the academic year of the absence. The calculations attribute all working days lost to the period the startdate falls into. 

The number of schools in our dataset varies year on year, which is why we have drawn on the percentage of days lost rather than number. Below you will find the number of schools in each year’s dataset. 

18/19 – 3,385

19/20 – 3,424

20/21 – 3,503

21/22 – 3,611

22/3 – 3,631

This represents approximately 10% of the schools in the UK. 

At Arbor, we’re on a mission to help schools make the most out of their data. Not yet using our MIS? Find out more here.